[PATCH] Add CONFIG for -mregparm=3

From: Andi Kleen <ak@muc.de>, me.

Using -mregparm=3 shrinks the kernel further:

(compiled with gcc 3.4, without -funit-at-a-time, using the later and
together with -Os shrinks .text even more, making over 700KB difference)

4129346  708629  207240 5045215  4cfbdf vmlinux
3892905  708629  207240 4808774  496046 vmlinux-regparm

This one helps even more, >236KB .text difference. Clearly worth
the effort.

This patch adds an option to use -mregparm=3 while compiling the kernel.  I
did an LTP run and it showed no additional failures over an non regparm
kernel.

According to some gcc developers it should be safe to use in all gccs that
are still supports (2.95 and up)

I didn't make it the default because it will break all binary only modules
(although they can be fixed by adding a wrapper that calls them with
"asmlinkage").  Actually it may be a good idea to make this default with
2.7.1 or somesuch.

We add new kbuild infrastructure: the command

	scripts/gcc-version.sh $(CC)

will print out the version of gcc in a canonical 4-digit form suitable for
performing numerical tests against.
DESC
arch/i386/Makefile,scripts/gcc-version.sh,Makefile small fixes
EDESC
From: Serge Belyshev <33554432@mtu-net.ru>

arch/i386/Makefile:
*  omitted $(KBUILD_SRC)/ in script call.

scripts/gcc-version.sh:
*  GNU tail no longer supports 'tail -1' syntax.

We should consider adding -fweb option:

   vanilla:
   $ size vmlinux
      text    data     bss     dec     hex filename
   3056270  526780  386056 3969106  3c9052 vmlinux

   with -fweb:
   $ size vmlinux
      text    data     bss     dec     hex filename
   3049523  526780  386056 3962359  3c75f7 vmlinux

   Also note 0.1 ... 1.0% speedup in various benchmarks.
   This option is not enabled by default at -O2 because it
   (like -fomit-frame-pointer) makes debugging impossible.

diff --git a/arch/i386/Kconfig b/arch/i386/Kconfig
index e61cddb589f9d5e6a10a3c7f41c2e238332ad347..454ec7d88050272b3a8cae3b87852c87aa839803 100644 (file)
--- a/arch/i386/Kconfig
+++ b/arch/i386/Kconfig
@@ -838,6 +838,19 @@ config BOOT_IOREMAP
        depends on (((X86_SUMMIT || X86_GENERICARCH) && NUMA) || (X86 && EFI))
        default y
 
+config REGPARM
+       bool "Use register arguments (EXPERIMENTAL)"
+       depends on EXPERIMENTAL
+       default n
+       help
+       Compile the kernel with -mregparm=3. This uses an different ABI
+       and passes the first three arguments of a function call in registers.
+       This will probably break binary only modules.
+
+       This feature is only enabled for gcc-3.0 and later - earlier compilers
+       generate incorrect output with certain kernel constructs when
+       -mregparm=3 is used.
+
 endmenu
 
 

diff --git a/arch/i386/Makefile b/arch/i386/Makefile
index 335c566bb6911298bf610447304f49a1031ea474..82c38e1acc4d678a51e857cbd636653410f29060 100644 (file)
--- a/arch/i386/Makefile
+++ b/arch/i386/Makefile
@@ -51,6 +51,11 @@ cflags-$(CONFIG_MVIAC3_2)    += $(call check_gcc,-march=c3-2,-march=i686)
 # AMD Elan support
 cflags-$(CONFIG_X86_ELAN)      += -march=i486
 
+# -mregparm=3 works ok on gcc-3.0 and later
+#
+GCC_VERSION                    := $(shell $(CONFIG_SHELL) $(srctree)/scripts/gcc-version.sh $(CC))
+cflags-$(CONFIG_REGPARM)       += $(shell if [ $(GCC_VERSION) -ge 0300 ] ; then echo "-mregparm=3"; fi ;)
+
 CFLAGS += $(cflags-y)
 
 # Default subarch .c files

diff --git a/include/asm-i386/module.h b/include/asm-i386/module.h
index 14ea6eb3f298f89428a97a77097a88c4ed3d1c24..d8aabc88fa433a8f2e9b558266678e1d8f4cbe54 100644 (file)
--- a/include/asm-i386/module.h
+++ b/include/asm-i386/module.h
@@ -54,6 +54,12 @@ struct mod_arch_specific
 #error unknown processor family
 #endif
 
-#define MODULE_ARCH_VERMAGIC MODULE_PROC_FAMILY
+#ifdef CONFIG_REGPARM
+#define MODULE_REGPARM "REGPARM "
+#else
+#define MODULE_REGPARM ""
+#endif
+
+#define MODULE_ARCH_VERMAGIC MODULE_PROC_FAMILY MODULE_REGPARM
 
 #endif /* _ASM_I386_MODULE_H */

diff --git a/scripts/gcc-version.sh b/scripts/gcc-version.sh
new file mode 100644 (file)
index 0000000..bb4fbea
--- /dev/null
+++ b/scripts/gcc-version.sh
@@ -0,0 +1,14 @@
+#!/bin/sh
+#
+# gcc-version gcc-command
+#
+# Prints the gcc version of `gcc-command' in a canonical 4-digit form
+# such as `0295' for gcc-2.95, `0303' for gcc-3.3, etc.
+#
+
+compiler="$*"
+
+MAJOR=$(echo __GNUC__ | $compiler -E -xc - | tail -n 1)
+MINOR=$(echo __GNUC_MINOR__ | $compiler -E -xc - | tail -n 1)
+printf "%02d%02d\\n" $MAJOR $MINOR
+